The present application relates generally to an improved data processing apparatus and method and more specifically to mechanisms for creating tag clouds based on user specified arbitrary shape tags.
Increasingly, people create textual “tags” and associate them with various forms of content, such as web pages and the like. These tags may be keywords indicative of the content and are themselves searchable, thereby enabling other users to find the associated content and retrieve the content for additional processing, such as display or analysis.
Textual tags may be rendered as a “tag cloud.” A tag cloud is a visual representation of the textual tags in which characteristics of the textual string of the textual tag, such as size or color, is set for each textual tag to a setting that represents the number of instances of that particular tag in a collection of tags. That is, for example, textual tags having a number of instances above a first threshold may have their textual string rendered with a first color or size in the tag cloud while other textual tags having a number of instances that are below this first threshold may have their textual string rendered in a second color or size in the tag cloud. In this way, a user can visualize which textual tags have a relatively larger number of instances than others. In other words, the various representations of the textual strings of the textual tags in the tag cloud may present a visualization of the relative popularity of each of the textual tags. An example mechanism for generating a tag cloud is described in commonly owned U.S. Pat. No. 7,752,534.
The use of tag clouds has recently been extended to image documents as well, so called “image clouds.” With such implementations, the size of the image in the visual array of the tag cloud is an indicator of the frequency of the image in, for example, a collection of web pages. The frequency of the image may be measured in terms of the number of times that the image is linked to in webpages on the Internet, for example. However, in the case of image clouds, the representations in the tag cloud, i.e. the tags, are thumbnail versions of already embedded images in web pages. There is no ability for an end user to specify what constitutes a tag to be used in the generation of an image cloud. The tags are the already embedded or linked to images, as a whole, in the web pages.